Caprine arthritis–encephalitis virus induces apoptosis in infected cellsin vitro through the intrinsic pathway
Angela Rea-Boutrois a, Guillemette Pontini b, Tim Greenland a, Patrick Mehlen c,Yahia Chebloune d, Gérard Verdier a, Catherine Legras-Lachuer a,⁎
a Université de Lyon, INRA, UMR754, Université Lyon 1, Ecole Nationale Vétérinaire de Lyon, Ecole Pratique des Hautes Etudes,IFR 128, 50 avenue Tony Garnier, 69 366 Lyon cedex 07, France
b INSERM U851, Université Lyon 1, IFR128-Biosciences Gerland, 21 avenue Tony Garnier, 69 007 Lyon cedex 07, Francec Apoptosis, Cancer and Development Laboratory, Laboratoire labellisé La Ligue, CNRS FRE2870, Centre Léon Bérard, 69008 Lyon, France
d The Kansas University of Medical Center, MMD Laboratory of Viral Pathogenesis, Kansas City, KS 66160, USA
Received 9 November 2007; returned to author for revision 13 December 2007; accepted 14 January 2008Available online 20 March 2008
The lentivirus genus of the retroviridae comprises complexnon-oncogenic viruses that naturally infect a range of mam-malian hosts, including humans (HIV), non-human primates(SIV), cattle (BIV), cats (FIV) horses (EIAV) and smallruminants (SRLV). The SRLV, CAEV and Maedi–Visna virus(MVV), originally found in goat and sheep respectively, causepersistent inflammatory lesions of the joints, the lungs, theudder and central nervous system of infected animals. TheSRLV infect cells of the monocyte/macrophage lineage, andviral replication in vivo is associated with the differentiation ofmonocytes into macrophages (Narayan et al., 1982; Zink et al.,1990). The SRLV do not infect CD4+ lymphocytes, and do notcause immunodeficiency in the infected animals (Narayan andClements, 1989).
Apoptosis plays an essential part in the normal developmentand homeostasis of multicellular organisms and also contributesto the elimination of damaged or virus-infected cells. Cell deathby apoptosis is characterized by specific morphological andbiochemical changes including cell shrinkage, nuclear chroma-tin condensation, DNA fragmentation, and proteolysis of highlyconserved cellular proteins by members of a cysteine-proteasesfamily called the caspases (Shi, 2002; Thornberry, 1997). Apo-ptosis may be induced either by the extrinsic pathway involvingcell surface death receptors or by the intrinsic pathway inducedby intracellular stimuli that transmit a signal to the mitochondria(Brenner and Kroemer, 2000). The extrinsic pathway ismediated by interaction of ligand with death receptors thattriggers the formation of multimeric complex followed by therecruitment and activation of caspase-8 (Lavrik et al., 2005).The intrinsic pathway involves an alteration of the mitochon-drial membrane potential (ΔΨ) leading to mitochondrialmembrane permeabilisation (MMP), and followed by a releaseof cytochrome c (Chipuk et al., 2006; Garrido et al., 2006).Cytolosic cytochrome c binds the protein adaptor APAF-1 to
453A. Rea-Boutrois et al. / Virology 375 (2008) 452–463
trigger the formation of a multimeric protein complex termedthe apoptosome that permits recruitment and autocleavage ofcaspase-9 (Bao and Shi, 2007). MMP is tightly regulated byBcl-2 family members, including Bcl-2 and Bax, that inhibit orpromote MMP respectively (Reed, 2006). Both the extrinsicand intrinsic pathways, converge downstream on the execu-tioner caspase-3, whose activity produces the morphologicalhallmarks of apoptosis (Porter and Janicke, 1999).
Apoptosis is an essential defense mechanism against invadingpathogens such as viruses or bacteria, and many viruses haveevolved strategies to control host apoptotic pathways at variousstages. However, apoptosis may also provide an importantmechanism for the release and spread of virus by reducing thehost's inflammatory or immune responses, and it can participate inthe development of viral pathogenesis (Barber, 2001; Benedictet al., 2002). The lentivirus HIV can induce apoptosis of bothinfected and uninfected CD4 T cells and this may be an importantmechanism for the depletion of CD4 T cells during AIDS (Hurtrelet al., 2005; Ross, 2001). Apoptosis of HIV-infected cells involvesboth the major pathways: the extrinsic pathway being mediatedby several death receptors including FAS or TRAIL receptorsand the intrinsic pathway being mediated by viral proteins thattarget mitochondria (Arnoult et al., 2004; Petit et al., 2003; Selliahand Finkel, 2001). Several viral proteins have been shown tobe involved in HIV-induced apoptosis (Varbanov et al., 2006)including Tat (Giacca, 2005), Nef (Zauli et al., 1999; Bouzar et al.,2007), Vpu (Akari et al., 2001), Vpr (Bouzar et al., 2004; Moonand Yang, 2006) and the envelope glycoprotein complex (gp120-gp41) (HolmandGabuzda, 2005; Perfettini et al., 2005). The SRLVcan also induce apoptosis in vitro infected cells. MVVacts (Duvalet al., 2002b) by both intrinsic and extrinsic pathways (Bellet et al.,2004; Duval et al., 2002a). For CAEV, it has been reported that invitro CAEV infection of goat synovial membrane (GSM) cells, athigh multiplicity of infection (m.o.i. of 10–20), is associated withmorphological changes characteristic of apoptosis, such aschromatin condensation and nuclear fragmentation (Gendelmanet al., 1997). However, the nature of apoptotic pathways involved inCAEV-infected cells has not previously been elucidated.
In the present study, we have investigated the apoptoticmechanism induced by CAEV infection of several caprine celltypes including primary macrophages, fibroblastic and epithelialcell lines. We show that infection of cells by CAEVeven at a lowmultiplicity (m.o.i. of 0.1) induces a significant increase ofapoptosis. Using specific caspase-9, -8 and -3 substrates as wellas dominant negative caspase-9, -8 and -3 expression constructs,we demonstrated that CAEV apoptosis uses the caspase-9pathway and coincides with a breakdown of the mitochondrialtransmembrane potential. Theses results have important impli-cations for our understanding of CAEV-induced apoptosis.
Results
Low doses of CAEVinduce apoptosis in goat synovial membrane(GSM) cells
It has previously been shown that infection of GSM cells byCAEV at high multiplicity of infection (m.o.i. of 10–20) pro-
vokes morphological changes characteristic of apoptosis (Gen-delman et al., 1997). We show here that CAEVinduces apoptosisin GSM cells at a more physiological level of infection. Apo-ptosis wasmeasured using theApostainmethod at different timespost-infection in GSM cells infected with CAEV-pBSCA or withthe CAEV-3112 isolate at a m.o.i. of 0.1. Mock-infected andstaurosporine treated GSM cells served as negative and positivecontrols.
We observed a progressive increase of apoptotic cells in GSMcell cultures infectedwithCAEV-pBSCAorwith the CAEV-3112isolate in comparison to mock-infected cells (Fig. 1). Increasedapoptosis was measurable at 3 days post-infection and increasedsignificantly by the sixth day post-infection, reaching 16% inCAEV-pBCSA infected cells, 19% in CAEV-3112-infected cellscompared to only 8% inmock-infected cells. The rate of apoptoticcells is 1.5 to 2.5-fold higher inCAEV-infectedGSMcells, than inmock-infected GSM cells, at 3, 5 and 6 days post-infection.
CAEV induces apoptosis in several caprine cell types
We examined the ability of CAEV to induce apoptosis in cellsother than GSM, including macrophages, epithelial and fibro-blastic cells. Caprine epithelial cells (TIGMEC) and primary goatmacrophages extracted from goat PBMC were infected withCAEV-pBSCA at a m.o.i. of 0.1. Immortalized caprine fibroblastcells (TIGEF), that are poorly susceptible toCAEVinfection, weretransfected with the pBSCA plasmid encoding the CAEV-CORKmolecular clone. Negative controls consisted of mock-infectedcells or cells transfectedwith pCMVplasmid and cells treatedwithstaurosporine served as positive controls. The number of cellsundergoing apoptosis was evaluated using the Apostain method,for TIGEFs cells (Fig. 2A) and TIGMEC cells (Fig. 2B). Primarygoatmacrophages cannot be analyzed by flow cytometry, soDAPIwas used to visualize cells with aberrant chromatin organizationand cells containing DNA strand breaks were evaluated by theterminal deoxynucleotidyl transferase (TdT)-mediated dUTP-biotin nick end-labelling (TUNEL) method (Fig. 2C).
As shown in Fig. 2, CAEV induced apoptosis in all cell lines.The rates of apoptosis increased significantly over time and,after 6 days, attained 18% in CAEV-infected TIGMEC cells,and 24% in transfected TIGEF cells. DAPI and TUNELanalysis of CAEV-infected primary macrophages showed theappearance of TUNEL- and DAPI-positive cells with morpho-logical changes characteristic of apoptosis, such as nuclearfragmentation and formation of apoptotic bodies. TUNEL-positive cells were counted from three independent experi-ments. At 6 days post-infection, 17% of macrophages infectedwith CAEV underwent apoptosis whereas only 10% of mock-infected macrophages were TUNEL positive.
These results demonstrate that the rate of apoptosis isapproximately 2 to 2.5 higher in CAEV-infected or transfectedcells than in negative control cells.
Apoptosis correlates with productive infection
We investigated the correlation between the proportion ofapoptotic cells and virus production by determining viral titers
Fig. 1. Analysis of apoptosis in infected GSM cells with CAEV-pBSCA. (A) GSM cells were inoculated with CAEV-pBSCA or CAEV-3112 at m.o.i of 0.1. At 3 and6 days post-infection, the rate of apoptosis was analyzed by flow cytometry following an incubation with a monoclonal antibody specific to single-stranded DNA, anda polyclonal anti-mouse IgM FITC conjugated antibody. Mock-infected cells were used as negative. In the case of staurosporine treatment, 2 µM of staurosporine wereincubated with GSM cells 24 h before analysis. The percentage of apoptotic cells was determined by measuring the percentage of those cells that were FITC positive(Apostain). (B) Kinetic of CAEV-induced apoptosis induction in GSM cells. CAEV-induced apoptosis in GSM cells was analyzed at different days post-infection.Data results are presented in histograms as the mean percentage of 5 independent experiments. Statistical differences between non infected cells and infected cells weredetermined by t-student test (⁎pb 0.05, ⁎⁎pb0.01).
454 A. Rea-Boutrois et al. / Virology 375 (2008) 452–463
over the time course of infection. Supernatants of previouslyinfected GSM cells, TIGMEC cells, and primary macrophagesas well as pBSCA-transfected TIGEF cells were harvested atdifferent times after infection. Titers of infectious cytopathicvirus were determined as described in material and methods.Virus titers increased progressively (Fig. 3A) up to 7 days post-infection when they reached 105 to 106 TCID50/ml forsupernatants from TIGMEC, TIGEF, macrophages and GSMcells. In addition, we analyzed the progression of infection inCAEV-infected GSM cells by immunocytochemistry using amonoclonal antibody specific to CAEV capsid p28 protein.Numeration of stained cells revealed an average of 30% and 70%
of p28-positive cells, at 3 days and 6 days post-infection,respectively (Fig. 3B). These results show that virus productionis correlated to the progression of CAEV infection and CAEV-induced apoptosis. However, the percentage of apoptotic cellsremained lower than that of infected cells.
To evaluate the proportion of cells that were both infectedand apoptotic, GSM cells were infected, then doubly stainedwith DAPI and TUNEL to visualize apoptotic cells, and byimmunocytochemistry using the p28 monoclonal antibody tovisualize infected cells. As shown in Fig. 3C, mock-infectedGSM cells, used as negative controls, showed a normal DAPInuclear staining and no evidence of p28 staining, CAEV-
Fig. 2. Analysis of CAEV-pBSCA-induced apoptosis in various caprine cells lines. TIGEF cells (A) were transfected with CAEV-pBSCA plasmid that encoded for thecomplete genome of CAEV-Cork strain or with the pCMV plasmid. TIGMEC cells (B) were infected with CAEV-pBSCAvirus at a m.o.i of 0.1. Mock-infected cells ortransfected cells without DNAwere used as negative control (NC). Analysis of apoptosis was performed by flow cytometry using Apostain method at different timespost-infection or post-transfection. Statistical differences between non-infected and infected cells were determined by t-student test (⁎pb0.05, ⁎⁎pb0.01,⁎⁎⁎pb0.001). (C) In situ detection of DNA fragmentation in primary goat macrophages. At 6 days post-infection, primary goat macrophages were stained usingTUNEL assay for apoptosis detection and counterstained with 0.5 µg/µl diamidinophenylindole (DAPI) for nuclei detection. TUNEL-positive cells were observed byfluorescent microscopy at magnification of ×400.
455A. Rea-Boutrois et al. / Virology 375 (2008) 452–463
Fig. 3. (A) Kinetic of CAEV virus production. Supernatants from CAEV-pBSCA infected cells or pBSCA-transfected cells were harvested on several days post-infection or transfection. Supernatants were titrated for infectious virus on GSM cells by using the Reed-Muench method, as described in materials and methods.(B) Immunofluorescence staining for CAEV p28 protein (gag). At 0, 3 and 6 days post-infection, CAEV-pBSCA infected GSM cells were stained byimmunofluorescence for detection of CAEV gag protein (p28) using a monoclonal antibody p28 and a polyclonal anti-mouse IgM FITC conjugated antibody. GSMcells were counterstained with DAPI (0.5 µg/µl) for nuclei detection prior microscopy observation (magnification×400). (C) Intracellular co-staining p28 and TUNELin CAEV-pBSCA infected cells. Mock or CAEV-pBSCA infected GSM cells were stained with TUNEL assay (red) and p28 monoclonal antibody (green) for co-detection of apoptotic cells and infected cells at 3 days post-infection. Before fluorescence microscopy observations, cells were counterstained with DAPI for nucleidetection. Panels show fluorescence staining for DAPI, TUNEL and p28 and a combine color photograph in the same area.
456 A. Rea-Boutrois et al. / Virology 375 (2008) 452–463
infected cultures showed cells with typical apoptotic nuclearmorphology (condensed and fragmented nuclei), and cellspositive for both p28 and TUNEL staining. Both individualinfected cells and multinucleated syncytia exhibited positiveTUNEL staining. To estimate the proportion of DAPI-TUNEL-p28 positive cells, two hundred cells were counted
in three independent experiments. Results showed that, where-as only 25% of p28-positive cells are TUNEL positive (Fig. 3C),all TUNEL-positive cells are p28-positive. These resultssuggest that the apoptotic cells had been infected but thatrelatively few infected cells underwent apoptosis at a giventime.
Fig. 4. Caspase-3, -8 and -9 activities in CAEV-pBSCA infected cells. TIGMECcells were infected with CAEV-pBSCA or with filtered medium of uninfectedcells (mock-infected control). In the case of staurosporine treatment, 2 µMstaurosporine were incubated with TIGMEC cells 24 h before analysis. Atdifferent times post-infection, cell lysates were prepared and caspase-3, -8 and -9activities were analyzed using Ac-DEVD-AMC, Ac-IETD-AMC and Ac-LEHD-AMC fluorogenic substrates, respectively. Caspases activities are expressed asfold change of caspase activity that is the ratio between the fluorescence obtainedin the CAEV-infected cells and the fluorescence measured in non-infected cells.The histogram presented the mean of three independent experiments. Statisticaldifferences between non infected cells and infected cells were determined byt-student test (⁎pb0.05, ⁎⁎pb0.01).
Fig. 5. Effect of dominant negative proteins of caspases-3 -8 and -9 in CAEV-induced apoptosis. (A) Human dominant negative proteins of caspases-3, -8 and -9block apoptosis in caprine cells. Caprine TIGMEC cells were transfected withpCMV (negative control) or with pcDNA-DNcasp8 or pcDNA-DNAcasp9 orpCDNA-DNcasp3 that encoded respectively the dominant negative proteins ofcaspases-8, -9 and -3. At 1 day post-transfection, transfected-cells were treatedwith staurosporine (S) and after a further incubation of 24 h, the rate of apoptoticcells were analyzed by flow cytometry following Apostain staining. TIGMECcells transfected with pCMV were used as negative control. Data results arepresented as the fold change of apoptosis induction that is the ratio between thepercentage of apoptotic cells obtained in each assay and the percentage ofapoptotic cells measured in control cells. (B) Analysis of CAEV-induced apoptosisin TIGMEC cells. TIGMEC cells infected with CAEV-pBSCA at a m.o.i 0.1 weretransfected with pcDNA-DNcasp8, pcDNA-DNAcasp9, and pcDNA-DNcasp3 orwith pCMVat 1 day post-infection.Mock-infectedTIGMECcells were transfectedwith pCMVand used as negative control (NC). At 3 and 6 days post-infection, therate of apoptotic cells was analyzed by flow cytometry following Apostainstaining. The histograms represent the fold change of apoptosis induction that is theratio between the percentage of apoptotic cells obtained in each assay and thepercentage of apoptotic cells measured in control cells. Student's test was used forstatistical comparison of means values (⁎pb0.05, ⁎⁎pb0.01, ⁎⁎⁎pb0.001).
457A. Rea-Boutrois et al. / Virology 375 (2008) 452–463
CAEV-induced apoptosis involves caspase-9
To determine the involvement of the intrinsic or extrinsicapoptotic pathways in CAEV infection, we examined the
activation of caspases-8, -9 and -3. For these studies, we usedthe caprine TIGMEC cells that were easily transfectable andsusceptible to CAEV infection. TIGMEC cells were infectedwith CAEV-pBSCA at a m.o.i. of 0.1. In parallel, mock-infectedcells and cells treated with staurosporine were used as negativeand positive controls. At 2, 3 and 6 days post-infection, celllysates were subjected to a fluorescent assay with substrates
458 A. Rea-Boutrois et al. / Virology 375 (2008) 452–463
specific for caspase-3, -8 and -9 activities. As shown in Fig. 4,we detected significant caspase-3 and -9 activities. The activitiesof caspases-3 and -9 increased over time to reach maximal valueat 6 days post-infection. At this time-point, caspase-3 and -9activities were 2 to 2.6 fold higher in CAEV-infected TIGMECcells than in mock-infected cells (Fig. 4A, B). In contrast, nosignificant activation of caspase-8 was observed in CAEV-infected cells (Fig. 4C).
In addition, dominant negativemutants of caspases-3, -8 and -9were used to confirm the involvement of these caspases in CAEV-induced apoptosis.Dominant negative caspase-encodingplasmidscarrying mutations in the active site cysteine to alanine were used.
We first assessed whether these dominant negative caspase-encoding plasmids were able to block cell death in caprine cells.Caprine TIGMEC cells were transfected with pcDNA-DNcasp3,
Fig. 6. (A) In situ detection of mitochondrial transmembrane potential in CAEV-pTIGMEC cells infected with CAEV-pBSCA and mock-infected cells by using Mitolfluorescent microscopy with appropriate filter (original magnification x400). (B) Influinfection, CAEV-infected TIGMEC cells were transfected with pcDNA-Bcl2 or pCnegative control (NC). At 3 and 6 days post-infection, the DNA fragmentation was stacytometry. The histograms represent the fold change of apoptosis induction that is thpercentage of apoptotic cells measured in control cells. Student's test was used for
pcDNA-DNcasp8, and pcDNA-DNcasp9, which encode the do-minant negative mutants for caspases-3, -8 and -9, respectively.At 1 day post-transfection, cells were treated with staurosporineand, at 2 days post-transfection, cell death was measured byflow cytometry using Apostain method. As shown in Fig. 5A,an important inhibition of staurosporine-induced cell deathwas observed in the presence of dominant negative mutantsfor caspases-8, -3 and -9, showing that these dominant negativemutants function in caprine cells.
To assess whether caspase inhibition could block CAEV-induced apoptosis, CAEV-infected TIGMEC cells were tran-siently transfected with pcDNA-DNcasp3, pcDNA-DNcasp8, orpcDNA-DNcasp9. At 3 and 6 days post-infection, cell death wasmeasured by flow cytometry (Apostain). As shown in Fig. 5B,our results show a significant inhibition of CAEV cell death in
BSCA infected cells. Mitochondrial transmembrane potential was analyzed inight detection kit, at 2 and 3 days post-infection. Stained cells were observed byence of anti-apoptotic Bcl-2 protein in CAEV-induced apoptosis. At 1 day post-MV. Mock-infected TIGMEC cells were transfected with pCMV and used asined using Apostain method and the rate of apoptotic cells was analyzed by flowe ratio between the percentage of apoptotic cells obtained in each assay and thestatistical comparison of means values (⁎pb0.05, ⁎⁎pb0.01, ⁎⁎⁎pb0.001).
459A. Rea-Boutrois et al. / Virology 375 (2008) 452–463
the presence of dominant negative mutants for caspase-9, and acomplete inhibition of apoptosis in the presence of dominantnegative mutants for caspase-3, suggesting that caspases-3 and-9 are required for induction of CAEV cell death. However, nosignificant inhibition was observed in the presence of dominantnegative proteins for caspase-8.
In addition, we assessed the caspase-3, -8 and -9 activities intransfected cells as for non-transfected cells at 3 days post-infection with CAEV-pBSCA. Specific inhibition of the relevantcaspase was observed in all cases (data not shown).
Taken together, these results clearly demonstrate CAEV-induced apoptosis involves the caspase-9 pathway.
CAEV-induced apoptosis depends on the mitochondrial pathway
Apoptotic signal transduction via the caspase-9 pathway isinitiated bymitochondrial damage that triggers a disruption of themitochondrial transmembrane potential (ΔΨ) resulting inmembrane permeabilisation and the release of apoptogenic factorsuch as cytochrome c from the mitochondria. Measurements ofΔΨ disruption can therefore provide information about theupstream apoptotic events. We therefore, used the Mitolightlipophilic cationic agent (5,5′,6,6′-tetrachloro-1,1′,3,3′-tetra-ethylbenzimidazolyl carbocyanine iodide) to evaluate changesin ΔΨ of infected cells. In healthy cells, MitoLight reagentaccumulates and aggregates in mitochondria, where it emits a redfluorescence. In apoptotic cells with altered ΔΨ, the dye in itsmonomeric form remains in the cytoplasm and emits a greenfluorescence. The possible disruption of ΔΨ during CAEV-induced apoptosis was investigated in TIGMEC cells infectedwith CAEVat a m.o.i of 0.1. At 2 and 3 days post-infection, cellswere incubated in theMitolight reagent for 30min, as described inmaterials and methods, and observed under the microscope. Asshown in Fig. 6A, mock-infected cells showed orange-red spots,indicating healthymitochondria with an aggregated phenotype. Ahigh percentage of CAEV-infected cells showed green fluores-cence, characteristic of an altered ΔΨ. This finding indicates adisruption of the mitochondrial membrane potential and suggeststhe involvement of mitochondria in CAEV-induced apoptosis.
Bcl-2 is an anti-apoptotic protein that prevents mitochondrialmembrane permeabilisation (Hockenbery et al., 1993). Wetherefore analyzed CAEV-induced apoptosis in cells that over-express Bcl-2 protein. CAEV-infected TIGMEC cells weretransfected with pcDNA-Bcl2 and the rate of apoptosis wasmeasured at 3 and 6 days post-infection by flow cytometryusing the Apostain method.
Results (Fig. 6B) show that over-expression Bcl-2 issufficient to inhibit CAEV-induced cell death. These resultsclearly demonstrate that CAEV induces apoptosis in caprinecells using the mitochondrial pathway.
Discussion
Over the past few years, a growing number of DNA and RNAviruses have been found to induce apoptosis in host cells (Clarkeand Clem, 2003; Nagaleekar et al., 2007; St-Louis andArchambault, 2007; Summerfield et al., 2001). Induction of
apoptosis following primate lentiviral infection (HIVand SIV) iswell documented (Hurtrel et al., 2005; Ross, 2001) and isconsidered to play a critical role in development of AIDS(Alimonti et al., 2003; Gougeon, 2005).Other lentiviral infectionshave also been shown to be associatedwith apoptosis notably FIV(Holznagel et al., 1998) and BIV (Xuan et al., 2007). For theSRLV, it has been demonstrated that in vitro MVV infectioninduced apoptosis in infected cells (Duval et al., 2002b) by boththe intrinsic and extrinsic pathways (Bellet et al., 2004; Duvalet al., 2002a). A previous study reported that in vitro CAEVinfection at high multiplicity of infection was associated withinduction of apoptosis in Tahr cells and GSM cells (Gendelmanet al., 1997), but the involvement of the intrinsic or extrinsicapoptotic pathways and the activation of caspases in CAEV-infected cells have not been investigated previously.
In the present study, we find that infection at a low m.o.i. ortransfection with CAEV results in increased apoptosis in severalcaprine cells including milk epithelial cells, synovial membranecells, immortalized goat embryo fibroblasts and primary caprinemacrophages isolated from blood. Similar results (data notshown) were also obtained using immortalized ovine macro-phages, MOCL4 cells line (Olivier et al., 2001), transfected withCAEV (they are remarkably resistant to direct infection). Overtime the rate of apoptosis increased concomitantly with theincrease in number of infected cells. However even when some60 to 80% of cells were positive for a viral antigen, indicatingproductive infection, only 17 to 25% of cells were apoptotic.Double labeling of CAEV capsid protein (p28) and DNAfragmentation (TUNEL assay) confirmed that, whereas allapoptotic cells in the infected cultures at 6 days after infectionexpressed viral antigen, only some 25% of the virally infectedcells showed signs of apoptosis by the DAPI or TUNELmethods. Whereas apoptosis after viral infection is usuallyregarded as a cellular attempt to limit viral replication, in somecases it might be a requirement for efficient virus assembly orrelease (Barber, 2001; Benedict et al., 2002).
Apoptosis may be induced by the extrinsic pathway whichactivates caspase-8 through cell receptors, principally, or theintrinsic pathway, where caspase-9 is activated via involvementof the mitochondria. In both cases, the result is the activation ofthe executioner caspase-3, which mediates the consequent lethalchain of events (Brenner and Kroemer, 2000; Porter and Janicke,1999). We confirmed the role of caspase-3 in CAEV-infectedcells by measuring an increase in the fluorescence in thepresence of a specific fluorogenic substrate in infected,compared to control cells. We also show that transfection ofthe target cells with a dominant negative variant of caspase-3abrogates the apoptotic response to infection by CAEV. Thisindicates that the induction of apoptosis by CAEV usesexclusively the caspase-dependant pathway. Fluorescenceassays indicate that caspase-9, but not caspase-8 is more activein the infected than the control cells, and the dominant negativevariant of caspase-9 reduces the rate of apoptosis after CAEVinfection by 60%. We also found a small inhibition (1.2 fold) inthe presence of the dominant negative variant of caspase-8, sothe extrinsic pathway might still play a minor part. In parallel,analysis of caspases-9, -8 and -3 in TIGMEC cells that express
460 A. Rea-Boutrois et al. / Virology 375 (2008) 452–463
dominant negative proteins show that caspase activities reflectthe level of apoptosis observed (data not shown). Like a previousstudy on the closely-related Maedi–Visna virus, we found anincrease caspases-3 and -9 in cells infected by CAEV. Incontrast, caspase-8, which was elevated in MVV-infected cells,remained at basal level in CAEV-infected cells. This mightrelated to the differences in cell type in the previous study, or todifference in methods for measuring caspase activities.
The evidence for involvement of the intrinsic pathway afterCAEVinfection led us to investigate a key event in the activationof caspase-9; mitochondrial membrane permeabilization. Thedisruption of the mitochondrial membrane potential allowsleakage into the cytoplasm of apoptogenic internal mitochon-drial components, such as cytochrome c (Chipuk et al., 2006;Garrido et al., 2006). We show that the mitochondrial membranepotential is indeed altered in cells infected by CAEV, and thatover-expression of Bcl-2, a protector of the mitochondrialmembrane, significantly reduces CAEV-induced apoptosis.Moreover, analysis of mitochondrial membrane potential orcaspase-9 and -3 activities in CAEV-infected TIGMEC cells thatexpress Bcl-2 proteins shows a normal mitochondrial membranepotential (data not shown) and no activation of caspases-3 and -9in these cells, clearly suggesting that Bcl-2 protein inhibitsCAEV-induced apoptosis. These results clearly indicate thatCAEV induces apoptosis in a range of host cells by the intrinsicpathway, permeablizing the mitochondrial membrane andcausing the release of cytochrome c with consequent activationof caspase-9 and subsequently, caspase-3.
It well established that SIV and HIV can induce apoptosisthrough both the intrinsic and extrinsic pathways. Not surpris-ingly, these viruses encode many proteins that can promoteapoptosis by targeting mitochondria or death receptor pathways,such as Vpr (Bouzar et al., 2004; Moon and Yang, 2006), Tat(Giacca, 2005), protease (Nie et al., 2007), Nef (Bouzar et al.,2007; Zauli et al., 1999) and the envelope glycoprotein complex(gp120-gp41) (Holm and Gabuzda, 2005; Perfettini et al., 2005).The SRLV viral protein(s) implied in apoptosis remainunknown, but we demonstrate here for the first time thatCAEV-induced apoptosis depends on activation of caspases-3and -9 and uses the mitochondrial pathway.
Materials and methods
Viruses and plasmids
CAEV-pBSCA virus was produced from the pBSCAplasmid, which encodes the complete genome of CAEV-CORK strain. The plasmid construct, and virus production havebeen described previously (Mselli-Lakhal et al., 1998). CAEV-3112, a French field isolate, was obtained from synovialmembrane explants from a naturally-infected goat (Blondinet al., 1989), and maintained in GSM cells.
The pcDNA-DNcasp3, pcDNA-DNcasp9 and pcDNA-DNcasp8 plasmids encode dominant negative proteins of humancaspases-3, -9 and -8 respectively under the control of CMVpromoter (Forcet et al., 2001) and pcDNA-Bcl2 expression vectorencodes full length human bcl-2 cDNA under the control of CMV
promoter (Hockenbery et al., 1993). Expression vector pBK-CMV(pCMV) was purchased from Stratagene (Stratagene, France).
Virus titration
Infectious virus was titrated on subconfluent monolayers ofindicator GSM cells inoculated with serial dilutions of filteredsupernatants from CAEV-infected or -transfected cells. Theinoculate was washed away after 4 h of contact with the cellsand replaced with fresh medium that was changed every 3 days.After 8 days of incubation, themonolayerswere stainedwithMay–GrünwaldGiemsa and examined for the presence of syncytia (giantmultinucleated cells). Viral titers were calculated using the Reed–Muench method and expressed as tissue culture infectious dose(TCID50) per millilitre of supernatant (Reed and Munch, 1938).
Cells
Goat synovial membrane (GSM) cells were originally obtainedfrom explanted carpal synovial membrane of a colostrum-deprived newborn goat as previously described (Narayan et al.,1980). The cells were cultured in Eagle's minimum essentialmedium (MEM; Gibco, Invitrogen France) supplemented with10% fetal bovine serum (FBS; hyclone, Perbio, France). Typically,cell monolayers were used for 4 to 8 passages.
Goat macrophages were derived from peripheral bloodmononuclear cells (PBMC) from blood of CAEV-negativegoats. Goat PBMCs were isolated by density-gradient centri-fugation using the Ficoll–Hypaque gradient, as previouslydescribed in (Bouzar et al., 2004; Bouzar et al., 2007). To obtainmature macrophages, 5×106 PBMC were cultured in macro-phage differentiation medium (MDM) supplemented with 20%heat inactivated lamb serum (ICN biomedicals, Orsay, France),and maintained 8 days in Teflon flasks at 37 °C in a 5% CO2.
T-immortalized goat milk epithelial cell (TIGMEC cells) is acontinuous cell line that was previously immortalized with Tantigen of SV40 (Mselli-Lakhal et al., 1999). Cells weremaintained in Dulbecco's modified eagle's medium (DMEM;Gibco, Invitrogen France) supplemented with 10% FBS.
Large T immortalized goat embryo fibroblast (TIGEF cells)is a continuous cell line that was previously immortalized withT antigen of SV40 (Mselli-Lakhal et al., 2001). Cells weremaintained in MEM supplemented with 10% FBS.
Transfection of plasmid DNAs and virus infection
TIGEF cells were seeded into 6-wells cell culture plates at adensity of 2.105 or 1.105 per well, respectively and maintainedin culture for 24 h before transfection with exgen500(Euromedex, France), according to the manufacturer's recom-mendations. Briefly, 3 µg of pBSCA or pBK-CMV plasmidDNAs (transfection control) were diluted into 150 µl of NaCl150 mM prior to the addition of 15 µl of Exgen500 diluted in150 µl of NaCl 150 mM. After incubation for 10 min at roomtemperature, the mixture was innoculated to the culture wellcontaining a final volume of 1 ml of medium, and the cultureplate was centrifuged 2 min at 800 rpm. The culture media was
461A. Rea-Boutrois et al. / Virology 375 (2008) 452–463
replaced with fresh medium 3 h after the transfection and cellswere incubated at 37 °C for further analysis.
Goat macrophages, GSM cells or TIGMEC were seeded into6-wells cell culture plates at a density of 1.105 per well. The dayafter, cells were infected at a multiplicity of infection (m.o.i) of0.1 with CAEV-pBSCA or CAEV-3112 and culture media wasreplaced 12 h after infection. In parallel, non infected cellsinoculated with filtered culture medium of non-infected cellswere used as negative control.
DNA breaks
Apoptosis was detected using the Apostain and TUNELmethods: Apostain labeled monoclonal antibody (Mab) specificto single-strand DNA (ssDNA) (F7-26) (Abcys, France) wasused according to the standard manufacturer's instructions.Stained cells were analyzed in a FACScan flow cytometer(Becton & Dickinson) with a 488 nm argon ion laser and lysis IIanalysis software for a minimum of 10.000 events. Percentagesof apoptotic cells reported in each Figure were determined fromthe number of cells showing fluorescein labeled DNA breaks.Positive controls consisted of cells treated with a non-specificinducer of apoptosis; Staurosporine (1 µM; Sigma, La Verpillère,France), 24 h before analysis. In situ DNA fragmentation wasmeasured using a TUNEL assay with ApopTag® Red (Chemi-con, Millipore, France) according to the manufacturer's instruc-tions. Briefly, goat macrophages or GSM cells were cultured onglass slides and infected with CAEV-pBSCA at a m.o.i of 0.1.Three or 6 days later, DNA breaks were stained, and slides weremounted with DABCO medium containing 0.5 µg/µl DAPI (4′-6-Diamidino-2-phenylindole) for examination by fluorescentmicroscopy with appropriate filters.
Intracellular P28 staining
GSM cells and goat macrophages were grown to sub-confluence on glass coverslips placed in 24-well cell cultureplates, then infected with CAEV-pBSCA. At 3 and 6 days post-infection, infected and control slips were fixed by incubation for10 min at room temperature in 1% paraformaldehyde, andpermeabilazed by incubation for 5 min at −20 °C in Ethanol:Acetic acid (2:1). Viral antigens were detected by incubation for30 min at room temperature with mouse monoclonal antibodiesdirected against CAEV p28 (VMRD, Pullman,WA,USA) diluted1:500 in PBS-BSA 1%. Cells were then incubated for 30 min atroom temperature with a purified FITC-labelled goat-anti-mouseIgG purified antibody (Sigma, La Verpillère, France) diluted 1:40in PBS-BSA1%, rinsedwith 1× PBS, andmountedwithDABCOmounting-medium containing 0,5 µg/µl DAPI. Slides wereobserved by fluorescent microscopy with appropriate filters.
Assay of caspase activity
Ac-DEVD-AMC (Acetyl-Asp-Glu-Val-Asp-7-amino-4-methylcoumarin), Ac-IETD-AMC (Acetyl-Ile-Glu-Thr-Asp-7-amino-4-methylcoumarin), and ac-LEHD-AMC (Acetyl-Leu-Glu-His-Asp-7-amino-4-methylcoumarin) (Anaspec, Euromedex,
France) were used to detect activities of caspases-3, -9, and -8respectively. Briefly, 1×106 cells were homogenized in 100 µllysis buffer (10 mM HEPES pH 7.4, 2 mM EDTA, 0.1% 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate(CHAPS), 5 mM dithiothreitol (DTT), 1 mM phenylmethylsul-fonyl fluoride (PMSF), 10 μg/ml aprotinin, and 10 μg/mlleupeptin) and incubated for 20 min in ice. Cell lysates werecentrifuged at 14 000 ×g at 4 °C for 15 min, and then thesupernatants were assayed for protein concentration using theBradford protein assay. Fifty µM of fluorogenic peptidesubstrate: Ac-IETD-AMC, Ac-DEVD-AMC, or Ac-LEHD-AMC were incubated with 100 µg of total protein in reactionbuffer (50 mM PIPES, pH 7.4, 10 mM EDTA, 0.5% CHAPS,25 mM dithiothreitol) for 1 h at 37 °C. The fluorescence ofreleased AMC was monitored using a Victor station (Wallac) atan excitation wavelength of 355 nm and emission wavelength of460 nm. Index of caspase activity was calculated as the ratiobetween the fluorescence of released AMC in lysates frominfected cells and that of released AMC in lysates from non-infected cells. Staurosporine treated cells (1 µM) were used as apositive control, as above.
Inhibition of caspase activity
The activities of caspases-3, -9 and -8 were specificallyinhibited in CAEV-pBSCA infected TIGMEC by transfectionwith pcDNA-DNcasp3, pcDNA-DNcasp9, pcDNA-DNcasp8,and pCMV (negative control) by Exgen500 at 24 h post-infection. Two days after transfection, 5×105 transfected cellswere collected and apoptosis was devaluated by flow cytometryusing the Apostain method, as described in the previous section.
Changes in mitochondrial transmembrane potential occuringduring apoptosis induced by CAEV-pBSCA were examinedusing MitoLight® apoptosis detection kit (Chemicon, Millipore,France). Briefly, GSM cells and TIGMEC cells were grown onglass coverslips and then inoculated with CAEV-pBSCA at a m.o.i of 0.1. At 72 h and 144 h post-inoculation, infected and non-infected cells (negative control) were rinsed in PBS and thenincubated with MitoLight reagent for 20 min at 37 °C in a 5%CO2. Cells were resuspended in incubation buffer and analyzedby fluorescence microscopy using a band-pass filter (Fluor-escein and rhodamine).
Statistical analysis
The means of three to five independent experiments arepresented as results. Student's test was used for statistical com-parison of means values. P-values b0.05 were considered statis-tically significant.
Acknowledgments
The authors thank the Flow cytometry and PLATIMplatforms of IFR128 and Mrs Barbara Gineys and Marie-Pierre
462 A. Rea-Boutrois et al. / Virology 375 (2008) 452–463
Confort for the animal care and their technical assistance. Weacknowledge the French ministry of research for fellowships(A. Rea-Boutrois).
References
Akari, H., Bour, S., Kao, S., Adachi, A., Strebel, K., 2001. The humanimmunodeficiency virus type 1 accessory protein Vpu induces apoptosis bysuppressing the nuclear factor kappaB-dependent expression of antiapopto-tic factors. J. Exp. Med. 194 (9), 1299–1311.
Alimonti, J.B., Ball, T.B., Fowke, K.R., 2003. Mechanisms of CD4+ Tlymphocyte cell death in human immunodeficiency virus infection andAIDS. J. Gen. Virol. 84 (Pt 7), 1649–1661.
Benedict, C.A., Norris, P.S., Ware, C.F., 2002. To kill or be killed: viral evasionof apoptosis. Nat. Immunol. 3 (11), 1013–1018.
Blondin, I., Grillet, C., Thiogane, Y., 1989. Syncytia formation in cultures andanalysis of the protein composition of various strains of caprine arthritisencephalitis virus (CAEV). Ann. Rech. Vet. 20 (2), 153–158.
Chipuk, J.E., Bouchier-Hayes, L., Green, D.R., 2006. Mitochondrial outermembrane permeabilization during apoptosis: the innocent bystanderscenario. Cell Death Differ. 13 (8), 1396–1402.
Clarke, T.E., Clem, R.J., 2003. In vivo induction of apoptosis correlating withreduced infectivity during baculovirus infection. J. Virol. 77 (3), 2227–2232.
Duval, R., Bellet, V., Delebassee, S., Bosgiraud, C., 2002a. Implication ofcaspases during maedi–visna virus-induced apoptosis. J. Gen. Virol. 83 (Pt12), 3153–3161.
Duval, R., Delebassee, S., Cardot, P.J., Bosgiraud, C., 2002b. Visna virus-induced cytopathic effect in vitro is caused by apoptosis. Arch. Virol. 147(5), 943–959.
Forcet, C., Ye, X., Granger, L., Corset, V., Shin, H., Bredesen, D.E., Mehlen, P.,2001. The dependence receptor DCC (deleted in colorectal cancer) defines analternative mechanism for caspase activation. Proc. Natl. Acad. Sci. U. S. A.98 (6), 3416–3421.
Garrido, C., Galluzzi, L., Brunet, M., Puig, P.E., Didelot, C., Kroemer, G., 2006.Mechanisms of cytochrome c release from mitochondria. Cell Death Differ.13 (9), 1423–1433.
Gendelman, R., Orzech, Y., Mashiah, P., Birenbaum, M., Gazit, A., Yaniv, A.,1997. Productive replication of caprine arthritis–encephalitis virus isassociated with induction of apoptosis. J. Gen. Virol. 78 (Pt 4), 801–805.
Giacca, M., 2005. HIV-1 Tat, apoptosis and the mitochondria: a tubulin link?Retrovirology 2, 7.
Gougeon, M.L., 2005. To kill or be killed: how HIV exhausts the immunesystem. Cell. Death. Differ. 12. Suppl. 1, 845–854.
Hockenbery, D.M., Oltvai, Z.N., Yin, X.M., Milliman, C.L., Korsmeyer, S.J.,1993. Bcl-2 functions in an antioxidant pathway to prevent apoptosis. Cell75 (2), 241–251.
Holm, G.H., Gabuzda, D., 2005. Distinct mechanisms of CD4+ and CD8+ T-cellactivation and bystander apoptosis induced by human immunodeficiencyvirus type 1 virions. J. Virol. 79 (10), 6299–6311.
Holznagel, E., Hofmann-Lehmann, R., Leutenegger, C.M., Allenspach, K.,Huettner, S., Forster, U., Niederer, E., Joller, H., Willett, B.J., Hummel, U.,Rossi, G.L., Schupbach, J., Lutz, H., 1998. The role of in vitro-inducedlymphocyte apoptosis in feline immunodeficiency virus infection: correla-tion with different markers of disease progression. J. Virol. 72 (11),9025–9033.
Hurtrel, B., Petit, F., Arnoult, D., Muller-Trutwin, M., Silvestri, G., Estaquier, J.,2005. Apoptosis in SIV infection. Cell Death Differ 12 (Suppl 1), 979–990.
Lavrik, I., Golks, A., Krammer, P.H., 2005. Death receptor signaling. J. Cell Sci.118 (Pt 2), 265–267.
Moon, H.S., Yang, J.S., 2006. Role of HIV Vpr as a regulator of apoptosis andan effector on bystander cells. Mol. Cells 21 (1), 7–20.
Mselli-Lakhal, L., Favier, C., Da Silva Teixeira, M.F., Chettab, K., Legras, C.,Ronfort, C., Verdier, G., Mornex, J.F., Chebloune, Y., 1998. Defective RNApackaging is responsible for low transduction efficiency of CAEV-basedvectors. Arch. Virol. 143 (4), 681–695.
Mselli-Lakhal, L., Guiguen, F., Fornazero, C., Du, J., Favier, C., Durand, J.,Grezel, D., Balleydier, S., Mornex, J.F., Chebloune, Y., 1999. Goat milkepithelial cells are highly permissive to CAEV infection in vitro. Virology259 (1), 67–73.
Nagaleekar, V.K., Tiwari, A.K., Kataria, R.S., Bais, M.V., Ravindra, P.V.,Kumar, S., 2007. Bluetongue virus induces apoptosis in culturedmammalian cells by both caspase-dependent extrinsic and intrinsic apoptoticpathways. Arch. Virol. 152 (9), 1751–1756.
Narayan, O., Clements, J.E., 1989. Biology and pathogenesis of lentiviruses.J. Gen. Virol. 70 (Pt 7), 1617–1639.
Narayan, O., Clements, J.E., Strandberg, J.D., Cork, L.C., Griffin, D.E., 1980.Biological characterization of the virus causing leukoencephalitis andarthritis in goats. J. Gen. Virol. 50 (1), 69–79.
Narayan, O., Wolinsky, J.S., Clements, J.E., Strandberg, J.D., Griffin, D.E.,Cork, L.C., 1982. Slow virus replication: the role of macrophages in thepersistence and expression of visna viruses of sheep and goats. J. Gen. Virol.59 (Pt 2), 345–356.
Nie, Z., Bren, G.D., Vlahakis, S.R., Schimnich, A.A., Brenchley, J.M., Trushin,S.A.,Warren, S., Schnepple, D.J., Kovacs, C.M., Loutfy, M.R., Douek, D.C.,Badley, A.D., 2007. Human immunodeficiency virus type 1 protease cleavesprocaspase 8 in vivo. J. Virol. 81 (13), 6947–6956.
Olivier, M., Berthon, P., Chastang, J., Cordier, G., Lantier, F., 2001.Establishment and characterisation of ovine blood monocyte-derived celllines. Vet. Immunol. Immunopathol. 82 (3–4), 139–151.
Perfettini, J.L., Castedo, M., Roumier, T., Andreau, K., Nardacci, R., Piacentini,M., Kroemer, G., 2005. Mechanisms of apoptosis induction by the HIV-1envelope. Cell Death Differ. 12 (Suppl 1), 916–923.
Petit, F., Arnoult, D., Viollet, L., Estaquier, J., 2003. Intrinsic and extrinsic pathwayssignaling during HIV-1 mediated cell death. Biochimie 85 (8), 795–811.
Porter, A.G., Janicke, R.U., 1999. Emerging roles of caspase-3 in apoptosis. CellDeath Differ. 6 (2), 99–104.
Reed, L.J., Munch, H., 1938. A simple method of estimating fifty percentendpoints. Am. J. Hyg. 27, 493–497.
Ross, T.M., 2001. Using death to one's advantage: HIV modulation ofapoptosis. Leukemia 15 (3), 332–341.
Selliah, N., Finkel, T.H., 2001. Biochemical mechanisms of HIV induced T cellapoptosis. Cell Death Differ. 8 (2), 127–136.
Shi, Y., 2002. Mechanisms of caspase activation and inhibition during apoptosis.Mol. Cell 9 (3), 459–470.
St-Louis, M.C., Archambault, D., 2007. The equine arteritis virus inducesapoptosis via caspase-8 and mitochondria-dependent caspase-9 activation.Virology 367 (1), 147–155.
463A. Rea-Boutrois et al. / Virology 375 (2008) 452–463
Summerfield, A., Zingle, K., Inumaru, S., McCullough, K.C., 2001. Inductionof apoptosis in bone marrow neutrophil-lineage cells by classical swine fevervirus. J. Gen. Virol. 82 (Pt 6), 1309–1318.
Thornberry, N.A., 1997. The caspase family of cysteine proteases. Br. Med.Bull. 53 (3), 478–490.
Varbanov, M., Espert, L., Biard-Piechaczyk, M., 2006. Mechanisms of CD4 T-cell depletion triggered by HIV-1 viral proteins. AIDS Rev. 8 (4), 221–236.
Xuan, C., Qiao, W., Li, J., Peng, G., Liu, M., Chen, Q., Zhou, J., Geng, Y.,2007. BTat, a trans-acting regulatory protein, contributes to bovine
Zauli, G., Gibellini, D., Secchiero, P., Dutartre, H., Olive, D., Capitani, S.,Collette, Y., 1999. Human immunodeficiency virus type 1 Nef proteinsensitizes CD4(+) T lymphoid cells to apoptosis via functional upregulationof the CD95/CD95 ligand pathway. Blood 93 (3), 1000–1010.
Zink, M.C., Yager, J.A., Myers, J.D., 1990. Pathogenesis of caprine arthritisencephalitis virus. Cellular localization of viral transcripts in tissues ofinfected goats. Am. J. Pathol. 136 (4), 843–854.
